Hello Andreas and Andrius, A combined email: > It looks quite valuable to me. Another good catch! :-) > Is the source for this PDF included as well? Debian ftpmaster has the > policy that pdf without (say LaTeX) source is "binary without source". > So we need to nag upstream to provide the source or if they do not > provide it need to exclude the documentation (which can be easily done > via the Files-Excluded field in debian/copyright). The exclusion can be > compensated by adding a link to the online version in a file > debian/README.Debian. Oh that's interesting. It doesn't look like it. I could ask upstream. One of the upstream developers wrote to me as an introduction and welcoming the interest in the packaging of their software which is really nice. On the other hand, the PDF is a direct replication of the github wiki[1] seemingly generated via pandoc[2], so perhaps we could just reference that for now until I can obtain a source. > If you was able to compile it it looks promising. Just try to follow > my initial advise and push your resulting Git repository. Than we > will see. Sounds great. > Thanks a lot for your effort > Thanks a lot for your contribution to Debian! I'm glad to hear that > packaging difficulties didn't scare you, but sparked your enthusiasm > instead :) This is one way to expand my scope of interest and knowledge right before university :-). Hopefully, I can keep building my way up to DM & DD in the future. > Excellent choice. In addition to Andreas's comment, I want to remark > that the package's source seems to include a bunch of "third-party" > sources (mostly under lib/). As you've already seen in MindTheGap, > dependencies that are already in Debian should be used instead of > embedded ones (if possible). A quick glance under lib/ reveals > 'gzstream', which seems to be packaged as 'libgzstream-dev' in Debian, > therefore it's better to remove it from the source (with Files > -Excluded, as mentioned by Andreas). Other dependencies are not > packaged, so I guess it's fine to leave them as is for now. Just be > sure to mention them in debian/copyright file (I usually start from > 'licensecheck -r --deb-machine'). All noted, thanks! [1]: https://github.com/soedinglab/mmseqs2/wiki [2]: https://pandoc.org/ On 15/07/2019 05:43, Andreas Tille wrote: > Hi Shayan, > > On Mon, Jul 15, 2019 at 01:48:52AM +0100, Shayan Doust wrote: >> Hi Andreas, >> >> Replying to this just in case you haven't seen it yet as I know you've >> probably been busy. Not sure if the initial message should also be >> targeted to the receipient email too or not. > > Pinging is fine. I was simply busy with real life. :-) > >> I assume you'll let me know whenever the MoM period is over too, as at >> the end of this month (from the 28th I think) I will be away for a week >> only. > > The period where you are free to ask here on the list is not limited. I > simply called that newcomer effort Mentoring of the Month to give some > realistic estimation how long a first packaging effort might last and to > make the student stick to the task with some patience. So just keep on > with your packaging effort here in public as long as you feel the need > to ask. > >> On 13/07/2019 15:58, Shayan Doust wrote: >>> Hello, >>> >>> Well, now that MindTheGap has been uploaded and will now rely on, of >>> course, long term support due to upstream updates now, I've been looking >>> at different softwares that have not yet been packaged. >>> >>> I stumbled across MMseqs2[1]. This is a search and clustering suite. I >>> have been unable to find this software within the debian repositories >>> for all releases so I can only assume this does not exist yet. I think >>> this categorises well under the general "life sciences" of things and >>> bioinformatics. What do you think? > > It looks quite valuable to me. Another good catch! > >>> I have also browsed through all the files, understood its directory >>> structure and have managed to compile it. This seems to be very well >>> documented too (fairly huge PDF is included). > > Is the source for this PDF included as well? Debian ftpmaster has the > policy that pdf without (say LaTeX) source is "binary without source". > So we need to nag upstream to provide the source or if they do not > provide it need to exclude the documentation (which can be easily done > via the Files-Excluded field in debian/copyright). The exclusion can be > compensated by adding a link to the online version in a file > debian/README.Debian. > >>> With enough time invested in this, would you say this is either too >>> huge/heavy to do, or do-able? > > If you was able to compile it it looks promising. Just try to follow > my initial advise and push your resulting Git repository. Than we will > see. > > Thanks a lot for your effort > > Andreas. > >>> [1]: https://github.com/soedinglab/MMseqs2 > >
Attachment:
signature.asc
Description: OpenPGP digital signature